We have developed a coaxial bioreactor that mimics the architecture of the liver asinus. It is composed of a polypropylene fiber (0.8 mm o.d.) centered inside an outer polypropylene fiber (1.8 mm i.d., 2.4 mm o.d) all contained inside a 5 mm NMR tube. Hepatocytes will be cultured in the space between the 2 fibers - the annular space. We want to mimic the physico-chemical gradients in the asinus and want to use MRM to measure radially across the annular space (1) perfusion and diffusion with and without hepatocytes, and (2) oxygen concentration using 19F MRM of perflourodecanoic acid. 1) The flow rates in the lumen of the inner tube (intracapillary space; ICS) and around the outer fiber (extracapillary space; ECS) will be varied with respect to each other, such that media will flow from the ECS through the annular space to the ICS. Then flow rates will be switch between ECS and ICS to visualize perfusion in the opposite direction. The difference in bulk flow between the ECS and ICS will be correlated with radial perfusion. Three images obtained using three ECS/ICS flow differentials will be obtained for each flow direction (i.e., ECS to ICS; ICS to ECS) - a total of 6 images. With hepatocytes in the annular space we believe coherent flow will be disrupted due to the resistance imposed by the tissue mass and diffusion will dominate. At a given ECS/ICS flow differential corresponding to a calibrated radial perfusion rate obtained from the experiment described above, we will measure the 'apparent' diffusion coefficient of water across the hepatocyte mass and/or measure radial perfusion at perhaps the fringe of the cell mass. 2) Oxygen concentration will be varied in the same manor that radial flow is varied by perfusing medium in the ECS and ICS that has been aerated with 40% and 20% oxygen plus air, respectively. T1-weighted 19F images will be obtained and will be calibrated to the intensity of the ECS where the oxygen concentration will be measured with an in-line electrode. The ECS intensity will need to be corrected for axial bulk flow effects. I will construct a 5 turn solenoid NMR probe that can be tuned to 1H and 19F. It will be wrapped around the outside of the 5mm NMR tube, so the FOV = 5 mm. The region of interest is only 0.5 mm (the annular space) and a voxel of 50 fm x 50 fm x 200 fm (will an anisotropic voxel effect relative quantitation?) will generate 10 voxels across the annular space. The absolute intensity of these 10 voxels will be plotted to generate 3 graphs: (1) ECS/ICS flow differential versus radial perfusion; (2) diffusion coefficient across the cell mass at one flow rate; (3) oxygen concentration across the cell mass at one flow rate. These data will be fit to determine function relationships.